Assessing the Validity of Self-Reported Adolescent Cigarette Smoking

Abstract

The comparison of adolescent cigarette smoking rates determined by traditional questionnaire, random response questionnaire and carbon monoxide test is reported. Results from 1,160 students in forty classes in seven different schools indicate the random response questionnaire elicited a statistically larger proportion of smokers than the traditional questionnaire. For practical purposes the magnitude of this difference may not warrant the added effort required to administer and interpret random response questionnaire results compared to carefully administered traditional questionnaires. Neither traditional questionnaire nor random response questionnaire results was greatly affected by the bogus pipeline technique.